How to take a picture of a black hole | Katie Bouman

譯者: Debra Liu 審譯者: Wilde Luo

In the movie "Interstellar,"

在「星際效應」這部電影中，

we get an up-close look at a supermassive black hole.

我們更近距離地看到了 超質量黑洞。

Set against a backdrop of bright gas,

在明亮氣體的背景下，

the black hole's massive gravitational pull

黑洞的巨大引力

bends light into a ring.

使光線形成戒指般的環狀。

However, this isn't a real photograph,

但是，這不是張真實的照片，

but a computer graphic rendering --

而是電腦圖像的呈現，

an artistic interpretation of what a black hole might look like.

是對黑洞可能的呈像的 藝術化的演繹。

A hundred years ago,

一百年前，

Albert Einstein first published his theory of general relativity.

愛因斯坦首先發表了他的相對論。

In the years since then,

那之後的幾年，

scientists have provided a lot of evidence in support of it.

科學家提供了很多證據支持他的理論。

But one thing predicted from this theory, black holes,

但是從他的理論中預測到的黑洞

still have not been directly observed.

仍然無法有直接證據證實。

Although we have some idea as to what a black hole might look like,

雖然我們對於黑洞的呈像有一些想法，

we've never actually taken a picture of one before.

但是我們從來沒有真正 拍攝過一張黑洞的相片。

However, you might be surprised to know that that may soon change.

也許你會驚訝於這種困境即將改變。

We may be seeing our first picture of a black hole in the next couple years.

我們在未來幾年內也許 可以得到第一張黑洞的相片。

Getting this first picture will come down to an international team of scientists,

國際的科學家團隊 將會獲得這第一張圖片，

an Earth-sized telescope

透過地球大小般的望遠鏡，

and an algorithm that puts together the final picture.

和一個演算方法，獲得最後這張圖片。

Although I won't be able to show you a real picture of a black hole today,

雖然，我今天無法讓大家看到 黑洞真正的照片，

I'd like to give you a brief glimpse into the effort involved

但是，我想要簡單地向各位說明一下

in getting that first picture.

獲得這首張照片所付出的努力。

My name is Katie Bouman,

我是 Katie Bouman ，

and I'm a PhD student at MIT.

一名麻省理工學院博士生。

I do research in a computer science lab

我在電腦科學實驗室做研究，

that works on making computers see through images and video.

讓電腦透過影像及影片， 能夠「看見」、識別。

But although I'm not an astronomer,

雖然我不是天文學家，

today I'd like to show you

但是，我現在要讓大家看的是

how I've been able to contribute to this exciting project.

我如何投入這令人興奮的專案。

If you go out past the bright city lights tonight,

如果今晚你們離開了城市明亮的燈光，

you may just be lucky enough to see a stunning view

可能運氣夠好，

可以看到銀河系美麗的影像。

of the Milky Way Galaxy.

And if you could zoom past millions of stars,

如果你的視野能夠穿越數百萬顆星星，

26,000 light-years toward the heart of the spiraling Milky Way,

向着銀河的螺旋中心前進 26,000 光年，

we'd eventually reach a cluster of stars right at the center.

最後會在中心點遇到一群星星。

Peering past all the galactic dust with infrared telescopes,

天文學家使用紅外線望遠鏡 透過銀河系塵埃

astronomers have watched these stars for over 16 years.

觀察這些星星， 已經超過了 16 年。

But it's what they don't see that is the most spectacular.

但是，最為壯觀的東西， 卻是他們無法看見的。

These stars seem to orbit an invisible object.

這些星星似乎繞著一個 隱形的物體運轉著。

By tracking the paths of these stars,

藉由追蹤這些星星的軌跡，

astronomers have concluded

天文學家得到一個結論：

that the only thing small and heavy enough to cause this motion

只有一個又小又重的物體 才能夠造成這樣的運動軌跡，

is a supermassive black hole --

那就是超質量黑洞，

an object so dense that it sucks up anything that ventures too close --

它的密度高到能夠吸收 所有敢於近距離靠近它的東西，

even light.

連光線也不例外。

But what happens if we were to zoom in even further?

但是，如果我們將影像放大， 會發生什麼事呢？

Is it possible to see something that, by definition, is impossible to see?

有沒有可能看到那些 原本被定義為看不見的東西呢？

Well, it turns out that if we were to zoom in at radio wavelengths,

事實顯示，如果我們 以無線電波長的尺度放大，

we'd expect to see a ring of light

我們預期可以看到一個光環，

caused by the gravitational lensing of hot plasma

它是由黑洞旁 高速移動的熱離子體的

zipping around the black hole.

「引力透鏡」效應形成。

In other words,

換句話說，

the black hole casts a shadow on this backdrop of bright material,

黑洞在明亮物質的背景下投射出陰影，

carving out a sphere of darkness.

刻畫出黑色的球體。

This bright ring reveals the black hole's event horizon,

這個光環揭露了黑洞的表面界限，

where the gravitational pull becomes so great

在那個地方，引力拉扯的力量很大，

that not even light can escape.

連光線都無法逃脫。

Einstein's equations predict the size and shape of this ring,

愛因斯坦方程式預測了 這個光環的大小與形狀，

so taking a picture of it wouldn't only be really cool,

所以拍攝黑洞的相片不只是很酷，

it would also help to verify that these equations hold

它也有助於驗證這些方程式

in the extreme conditions around the black hole.

能在黑洞附近這樣的 極端環境下成立。

However, this black hole is so far away from us,

但是，這個黑洞距離我們非常遙遠，

that from Earth, this ring appears incredibly small --

從地球看過去， 這個光環是不可思議的小，

the same size to us as an orange on the surface of the moon.

就像是月球表面的 一個橘子那樣的小。

That makes taking a picture of it extremely difficult.

所以拍攝黑洞的相片是極其困難的。

Why is that?

為什麼呢？

Well, it all comes down to a simple equation.

因為，這所有的一切 可以歸結於一個簡單的方程式。

Due to a phenomenon called diffraction,

由於「衍射」現象，

there are fundamental limits

我們所能觀察到的最小物體，

to the smallest objects that we can possibly see.

是有限小的， 我們無法洞察更小的結構。

This governing equation says that in order to see smaller and smaller,

這個方程式說， 為了要看到越來越小的物體，

we need to make our telescope bigger and bigger.

我們必須製作越來越大的望遠鏡。

But even with the most powerful optical telescopes here on Earth,

但是，即使透過地球上 最強大的光學望遠鏡，

we can't even get close to the resolution necessary

我們還是遠遠達不到

to image on the surface of the moon.

拍攝月球表面的影像所需要的解析度。

In fact, here I show one of the highest resolution images ever taken

事實上，請大家看看這張從地球拍攝的

解析度最高的月球照片之一，

of the moon from Earth.

It contains roughly 13,000 pixels,

這張相片大約有一萬三千像素，

and yet each pixel would contain over 1.5 million oranges.

而每一個像素可包含 超過 150 萬個橘子。

So how big of a telescope do we need

那麼，為了要看到月球表面的橘子，

in order to see an orange on the surface of the moon

我們需要多大的望遠鏡呢？

and, by extension, our black hole?

再者，為了要看到黑洞？

Well, it turns out that by crunching the numbers,

事實證明，透過大量運算，

you can easily calculate that we would need a telescope

我們可以很容易地計算出 我們所需要的望遠鏡

the size of the entire Earth.

必須是整個地球那麼大。

(Laughter)

(笑聲)

If we could build this Earth-sized telescope,

如果我們建造出地球般大小的望遠鏡，

we could just start to make out that distinctive ring of light

我們馬上就可以探測出一個獨特光環，

indicative of the black hole's event horizon.

它表明了黑洞的表面界限。

Although this picture wouldn't contain all the detail we see

雖然這張相片沒有包含所有細節，

in computer graphic renderings,

像我們在電腦圖形渲染上看到的那樣，

it would allow us to safely get our first glimpse

但是，至少我們可以安全地

of the immediate environment around a black hole.

對黑洞附近的環境瞥上一眼。

However, as you can imagine,

然而，如同大家想像的，

building a single-dish telescope the size of the Earth is impossible.

建造一個地球大小的 單碟望遠鏡是不可能的。

But in the famous words of Mick Jagger,

但是在 Mick Jagger 的名言中：

"You can't always get what you want,

「你無法一直得到你所想要的，

but if you try sometimes, you just might find

但是如果你去嘗試，你可能會發現

you get what you need."

你得到了你所需要的。」

And by connecting telescopes from around the world,

藉由連結世界各地的望遠鏡，

an international collaboration called the Event Horizon Telescope

名為「事件視界望遠鏡」的國際組織

is creating a computational telescope the size of the Earth,

正著手創建一個地球大小的 計算型望遠鏡，

capable of resolving structure

它能夠解析黑洞的

on the scale of a black hole's event horizon.

表面界限的結構。

This network of telescopes is scheduled to take its very first picture

這個望遠鏡網路預計明年

of a black hole next year.

拍攝黑洞的第一張相片。

Each telescope in the worldwide network works together.

世界各地的望遠鏡網路同時運作。

Linked through the precise timing of atomic clocks,

透過原子鐘的精準時間鏈結，

teams of researchers at each of the sites freeze light

每個地點的研究團隊 藉由蒐集數千兆兆字節的數據

by collecting thousands of terabytes of data.

將光線「定格」。

This data is then processed in a lab right here in Massachusetts.

麻薩諸塞州這裡的實驗室 接下來處理這些資料。

So how does this even work?

那麼，這些資料是如何運作的呢？

Remember if we want to see the black hole in the center of our galaxy,

還記得嗎？如果我們想要 看清在銀河中間的黑洞，

we need to build this impossibly large Earth-sized telescope?

我們就需要建造地球大小的 望遠鏡，這是不現實的。

For just a second, let's pretend we could build

等一下，假設我們能夠建造

a telescope the size of the Earth.

地球般大小的望遠鏡。

This would be a little bit like turning the Earth

就有點像將地球

into a giant spinning disco ball.

想像成舞廳裡的迪斯可旋轉球。

Each individual mirror would collect light

每一面鏡子會蒐集光線，

that we could then combine together to make a picture.

然後我們能將這些 影像整合成一張圖片。

However, now let's say we remove most of those mirrors

但是，現在讓我們 移除大多數的鏡子，

so only a few remained.

只剩下少數幾個。

We could still try to combine this information together,

我們仍可試著整合這些資訊，

but now there are a lot of holes.

但是，現在只能看到很多「孔洞」。

These remaining mirrors represent the locations where we have telescopes.

這些剩下的鏡子代表 那些有望遠鏡的地方。

This is an incredibly small number of measurements to make a picture from.

測量數據少之又少， 甚至無法形成一張圖片。

But although we only collect light at a few telescope locations,

雖然我們只在少數 有望遠鏡的地方蒐集光線，

as the Earth rotates, we get to see other new measurements.

地球旋轉時，我們可以 獲得一些新的測量數據。

In other words, as the disco ball spins, those mirrors change locations

換句話說，就像迪斯可球旋轉時， 那些鏡子也會改變位置，

and we get to observe different parts of the image.

我們得以觀察不同面向的影像。

The imaging algorithms we develop fill in the missing gaps of the disco ball

我們所開發的成像算法填補了 「迪斯可球」的不可見縫隙，

in order to reconstruct the underlying black hole image.

目的在重建黑洞的相片。

If we had telescopes located everywhere on the globe --

如果地球的每個地方都有望遠鏡，

in other words, the entire disco ball --

也就是整個迪斯可球佈滿了鏡子，

this would be trivial.

這是最簡潔、理想的情況。

However, we only see a few samples, and for that reason,

但是，我們只看得到 某些局部的成像，因此，

there are an infinite number of possible images

有無數可能的相片

that are perfectly consistent with our telescope measurements.

可以與現有望遠鏡的 局部成像相吻合。

However, not all images are created equal.

當然，並不是每一張「相片」的 優先級別都相同。

Some of those images look more like what we think of as images than others.

有些相片比別的 更近似我們所想像的。

And so, my role in helping to take the first image of a black hole

因此，為了協助拍攝黑洞 第一張相片，我的任務就是

is to design algorithms that find the most reasonable image

設計發現最合理影像的演算法，

that also fits the telescope measurements.

當然也必須符合望遠鏡的量測數據。

Just as a forensic sketch artist uses limited descriptions

就像法庭的素描家一樣， 利用有限的相貌描述以及

to piece together a picture using their knowledge of face structure,

他們對於臉部結構的知識， 將表現相貌特點的圖片拼湊出來，

the imaging algorithms I develop use our limited telescope data

我開發的影像演算法 使用有限的望遠鏡資料

to guide us to a picture that also looks like stuff in our universe.

為我們生成這種影像： 類似於宇宙中的事物的影像。

Using these algorithms, we're able to piece together pictures

利用這些演算法， 讓我們能夠利用零零散散的資料

from this sparse, noisy data.

拼湊出黑洞可能的樣子。

So here I show a sample reconstruction done using simulated data,

在這裡，讓大家看一個利用模擬資料 重建的影像樣本，

when we pretend to point our telescopes

這是我們假設將望遠鏡指向

to the black hole in the center of our galaxy.

銀河系中心的黑洞時所得到的。

Although this is just a simulation, reconstruction such as this give us hope

雖然這只是一個模擬， 但是這讓我們充滿了希望：

that we'll soon be able to reliably take the first image of a black hole

我們能夠仰賴這樣的模擬演算法， 很快地得到黑洞的第一張相片，

and from it, determine the size of its ring.

同時也能計算「光環」的大小。

Although I'd love to go on about all the details of this algorithm,

雖然我很樂意繼續說明 這個演算法的所有細節，

luckily for you, I don't have the time.

但由於時間不夠，所以 你們也不用費腦子聽了。

But I'd still like to give you a brief idea

但是，我還是很樂意 跟大家做個簡短的說明：

of how we define what our universe looks like,

我們如何定義宇宙看起來像什麼？

and how we use this to reconstruct and verify our results.

以及我們如何 利用這個演算法重建並驗證結果。

Since there are an infinite number of possible images

因為有無數可能的影像

that perfectly explain our telescope measurements,

與地球上望遠鏡的量測完全符合，

we have to choose between them in some way.

我們必須在它們之間 找個方法進行挑選。

We do this by ranking the images

我們對影像進行打分，

based upon how likely they are to be the black hole image,

打分的根據是：看起來有多像黑洞，

and then choosing the one that's most likely.

然後選擇最像的影像。

So what do I mean by this exactly?

那麼，這到底是什麼意思呢？

Let's say we were trying to make a model

假設我們試著建立一個模型，

that told us how likely an image were to appear on Facebook.

它告訴我們這個影像在 Facebook 上出現的可能性。

We'd probably want the model to say

我們希望這個模型會這樣判斷：

it's pretty unlikely that someone would post this noise image on the left,

大家應該不太可能會上傳 像左邊這張亂亂的圖，

and pretty likely that someone would post a selfie

而比較可能會上傳自拍照，

like this one on the right.

像右邊這張。

The image in the middle is blurry,

中間這張圖像片是模糊的，

so even though it's more likely we'd see it on Facebook

即使模糊，和亂亂的圖像比較的話， 我們還是很有可能

compared to the noise image,

在 Facebook 上看到，

it's probably less likely we'd see it compared to the selfie.

只不過不如自拍照那樣常見。

But when it comes to images from the black hole,

但是，如果是黑洞的影像，

we're posed with a real conundrum: we've never seen a black hole before.

我們遇到一個真正的難題： 我們從來沒見過黑洞的樣子。

In that case, what is a likely black hole image,

在這種情況下， 黑洞可能的影像是什麼？

and what should we assume about the structure of black holes?

我們應該假設黑洞的結構是什麼？

We could try to use images from simulations we've done,

我們可能會試著使用 之前生成的模擬結果，

like the image of the black hole from "Interstellar,"

像「星際效應」裡的黑洞影像，

but if we did this, it could cause some serious problems.

但是，如果這樣做的話， 可能會造成一些嚴重的問題。

What would happen if Einstein's theories didn't hold?

如果愛因斯坦的理論不適用的話， 會發生什麼事？

We'd still want to reconstruct an accurate picture of what was going on.

我們還是想要重建 一個準確的圖像。

If we bake Einstein's equations too much into our algorithms,

如果將太多愛因斯坦的方程式 融入我們的演算法中，

we'll just end up seeing what we expect to see.

最後只會得到我們期望的結果， 而不一定是事實。

In other words, we want to leave the option open

換句話說，我們不能 貿然確定實際情況如何，

for there being a giant elephant at the center of our galaxy.

因為銀河系中央有一隻巨象。

(Laughter)

(笑聲)

Different types of images have very distinct features.

不同類型的影像有著 各自非常顯著的特徵。

We can easily tell the difference between black hole simulation images

我們可以很容易地區分 黑洞模擬影像

and images we take every day here on Earth.

以及我們在地球上日常生活中的照片。

We need a way to tell our algorithms what images look like

我們需要一種方法來告訴演算法 影像看起來像什麼，

without imposing one type of image's features too much.

而不是去強加特定一種影像的特徵給它。

One way we can try to get around this

我們可以用一個方法 試著解決這個問題：

is by imposing the features of different kinds of images

通過導入不同類型影像的特徵 讓演算法重建影像，

and seeing how the type of image we assume affects our reconstructions.

然後觀察預先假設的影像類型 如何影響我們重建的影像。

If all images' types produce a very similar-looking image,

如果所有不同類型的影像特徵 產生的結果都很類似，

then we can start to become more confident

那麼我們可以充滿信心地說：

that the image assumptions we're making are not biasing this picture that much.

對於這個影像所做的假設 沒有與事實偏差太多。

This is a little bit like giving the same description

這有點像是將相同的相貌描述

to three different sketch artists from all around the world.

提供給三個來自世界各地不同的素描家，

If they all produce a very similar-looking face,

如果他們都畫出很相像的臉，

then we can start to become confident

那麼我們可以充滿信心地說：

that they're not imposing their own cultural biases on the drawings.

他們的作品沒有受到 本人的文化偏見的影響。

One way we can try to impose different image features

我們導入不同類型影像 的特徵的一個方法

is by using pieces of existing images.

就是藉由現存的影像去拼湊。

So we take a large collection of images,

所以我們要蒐集大量的影像，

and we break them down into their little image patches.

然後將它們分解成許多碎片。

We then can treat each image patch a little bit like pieces of a puzzle.

之後我們可以把這些碎片 當作拼圖的碎片。

And we use commonly seen puzzle pieces to piece together an image

我們使用常見的「碎片」拼湊成圖片，

that also fits our telescope measurements.

這張圖片當然也要 符合望遠鏡的量測數據。

Different types of images have very distinctive sets of puzzle pieces.

不同類型的影像有其獨特的拼圖碎片。

So what happens when we take the same data

所以，當我們利用相同的數據資料

but we use different sets of puzzle pieces to reconstruct the image?

卻使用不同類型的拼圖碎片 來重建這個影像，會發生什麼事？

Let's first start with black hole image simulation puzzle pieces.

讓我們先從黑洞模擬 圖像的拼圖碎片開始。

OK, this looks reasonable.

好的，這看起來很合理。

This looks like what we expect a black hole to look like.

這看起來像我們 所期待的黑洞的樣子。

But did we just get it

但是，僅僅是導入了 一些些黑洞模擬影像的碎片，

because we just fed it little pieces of black hole simulation images?

我們就得出了結果嗎？

Let's try another set of puzzle pieces

讓我們來試試另一組拼圖，

from astronomical, non-black hole objects.

這些是天文學影像的拼圖，不是黑洞的。

OK, we get a similar-looking image.

沒錯，我們得到一個類似的影像。

And then how about pieces from everyday images,

那麼如果是日常生活的影像呢？

like the images you take with your own personal camera?

就像用相機所照的照片一樣？

Great, we see the same image.

很好，我們得到相同的影像。

When we get the same image from all different sets of puzzle pieces,

當我們從不同類型的拼圖 得到相同的影像，

then we can start to become more confident

我們更有信心了，

that the image assumptions we're making

我們所假定的影像

aren't biasing the final image we get too much.

和我們最後得到的影像 並沒有差距太多。

Another thing we can do is take the same set of puzzle pieces,

我們可以做的另一件事 就是使用同一組拼圖，

such as the ones derived from everyday images,

比如日常生活中的影像碎片，

and use them to reconstruct many different kinds of source images.

並利用它們來重組 各種不同素材來源的影像。

So in our simulations,

那麼，在模擬實驗當中，

we pretend a black hole looks like astronomical non-black hole objects,

我們假設黑洞看起來就像是 天文學裡那些非黑洞的物體，

as well as everyday images like the elephant in the center of our galaxy.

或者又把它看成「銀河系中央的大象」 這樣的日常生活影像。

When the results of our algorithms on the bottom look very similar

我們下方的演算結果

to the simulation's truth image on top,

和上方的模擬實驗中的真實影像很相像，

then we can start to become more confident in our algorithms.

我們就可以對我們的演算法更有信心。

And I really want to emphasize here

我真的想要強調這一點：

that all of these pictures were created

這些所有的圖片都是

by piecing together little pieces of everyday photographs,

由日常生活照片的碎片 拼湊出來的，

like you'd take with your own personal camera.

就是那種用私人相機照出來的照片。

So an image of a black hole we've never seen before

我們之前從沒看過黑洞的相片，

may eventually be created by piecing together pictures we see all the time

但最後黑洞的相片也許是由我們 常常看到的日常生活照片拼湊出來的：

of people, buildings, trees, cats and dogs.

人像、建築物、樹木、貓、狗等等。

Imaging ideas like this will make it possible for us

這些成像方法讓我們能夠

to take our very first pictures of a black hole,

拍攝出黑洞的第一張相片，

and hopefully, verify those famous theories

我們同時也希望 能夠驗證那些著名的理論，

on which scientists rely on a daily basis.

那些科學家平常所依賴的理論。

But of course, getting imaging ideas like this working

當然，提出這些成像的方法與理論，

would never have been possible without the amazing team of researchers

沒有一個驚人的研究團隊 是不可能達到這種成果的，

that I have the privilege to work with.

我很榮幸身為這個團隊的一員。

It still amazes me

我對這件事感到驚異：

that although I began this project with no background in astrophysics,

雖然我沒有任何天文物理的背景 而加入這個專案，

what we have achieved through this unique collaboration

我們透過這獨特的合作所得到的，

could result in the very first images of a black hole.

能夠獲得第一張黑洞的相片。

But big projects like the Event Horizon Telescope

但是像「事件視界望遠鏡」 這樣的大專案，

are successful due to all the interdisciplinary expertise

多虧有跨學科領域的專業知識而成功，

different people bring to the table.

不同的專家共同合作着。

We're a melting pot of astronomers,

我們像是個熔爐，集結了天文學家、

physicists, mathematicians and engineers.

物理學家、數學家和工程師。

This is what will make it soon possible

這就是我們讓不可思議的事情

to achieve something once thought impossible.

快速實現的原因。

I'd like to encourage all of you to go out

我很想鼓勵大家

and help push the boundaries of science,

去協助推動科學的前沿，

even if it may at first seem as mysterious to you as a black hole.

即使第一步可能像黑洞那樣神秘。

Thank you.

謝謝大家。

(Applause)

(掌聲)